Why this question is really about incentives
Anyone commissioning a diagnostic whose fee is credited against the follow-on build has spotted the obvious problem: the firm is better off finding something. That is a fair objection and it does not have a rhetorical answer. It has a structural one, or it has none.
The structural answer has two parts. First, a written commitment that an immaterial finding is reported as immaterial, in the first paragraph, with an explicit recommendation not to proceed. Second, implementation priced and approved one phase at a time, with no committed total — because a firm that wanted to inflate findings would want a large committed build, not an approval it has to re-earn at every phase.
Ask for both in writing. A firm that will give you the credit but not the counterweight has built a sales funnel with a report attached to it.
The three shapes a disappointing result takes
A genuinely null result — no measurable leakage — is uncommon in an operation of any complexity. Far more common are three outcomes that all deserve the same honesty.
- 01Immaterial. Leakage exists and is small relative to the cost of closing it. A $40,000 annual leak against a $60,000 build is not a project, it is a rounding error with a project attached.
- 02Real but not operational. The loss is genuine and the cause is pricing, product or competitive — none of which an automation build addresses. The correct report says where the money is going and that this firm is the wrong instrument.
- 03Real but not closable yet. The leak is large and it sits behind something structural: a system migration already underway, a reorganisation, an acquisition mid-integration. Building now means building twice.
The third is the one most likely to be handled badly, because it is the one where a firm can technically deliver a build that technically closes a gap that is about to be rebuilt anyway. Nobody has lied and the money is wasted.
What you still own after a null result
Everything the engagement produced. The report and the roadmap are yours whether you build with the firm, build it yourself, or hand the whole thing to somebody else — and the roadmap is written so that another firm could execute it, which matters most precisely in the case where you are not going to use the firm that wrote it.
A null result also leaves you with something less obvious and quite valuable: a measured baseline of your own funnel that did not exist before. Latency by channel, conversion by latency, volume by source, the share of enquiries receiving no second contact. That baseline is what makes any future claim about operational improvement testable, and it is the thing companies most often discover they needed after it was no longer available.
What a null result tells you about the firm
More than a positive one does, which is the awkward part. Any firm can produce a report full of findings. A firm that has returned a null result and said so has demonstrated the only thing that makes its positive findings worth anything.
So it is a reasonable question to ask in a reference call, or of the firm directly: how many engagements have returned an immaterial finding, and what happened next. The answer being "none, ever" is not the reassurance it sounds like.
How to reduce the odds of paying for a null result
You can get most of the way to the answer yourself, for free, before commissioning anything. The point is not to pre-empt the diagnostic but to establish whether the order of magnitude justifies it.
- Split last quarter’s enquiries by whether they got a genuine human response inside an hour, and compare conversion between the two groups. A large gap with a large slow group is the signal.
- Count enquiries that received exactly one contact and no follow-up. If that number is small, one whole category of leak is already absent.
- Call your own business twice — mid-afternoon on a Tuesday, and on a Saturday evening. Note what happens and how long it takes. This costs ten minutes and is startlingly diagnostic.
- Check every channel you have ever published. The old inbox, the second number, the chat widget, the marketplace queue.
- Multiply the gap you found by your own average first-year value, and annualise it. If that figure is not comfortably a multiple of a $5,000 fee plus a $15,000 to $75,000 build, the honest answer is not yet.
If that rough arithmetic produces a large number, a diagnostic will tell you how much of it is real and in what order to attack it. If it produces a small one, you have your answer and it cost you an afternoon. Either way the work was worth doing, which is the useful property of a test that can come back negative.